Dataset: Dead mental health startups, 2000-2026, coded on 18 fields

30 points by trubetskov 21 days ago on hackernews | 20 comments

Shouldn't this be a private dataset?

Isn’t this parent comment generated by AI?

: whydoineedthis | 21 days ago
I know you are AI, but what am I?

Why? OP has assembled it out of public information via company filings etc., and is releasing it as a supplement to a report to the field. It seems like you didn't read the accompanying notes at all.

harvey9 | 21 days ago

Why?

orsenthil | 21 days ago

Mental Health Data.

harvey9 | 21 days ago

Are you serious? It's data about corporations, not your shrink's notes from your consult.

VCFundedGenYer | 21 days ago

HIPAA is a real thing. It is to be respected.

This study appears to be LLM generated so all confidence in it being a) accurate and b) respective of regulations has all gone out the window.

: [OP] trubetskov | 21 days ago
why so?
: harvey9 | 21 days ago
I'm not in America but just a guess, HIPAA does not apply to a company's financial filings. LLMs are beside the point here.

[OP] trubetskov | 21 days ago

why should it be so? we collected it from open sources and analised very carefully to put all the markers. so i think there is no law that can say "make it private".

palmotea | 21 days ago

Does this include the several stupid "thin wrapper on ChatGPT to 'help' people with depression/ADHD/etc." startups that I recall from about two seconds after ChatGPT came out?

Basically: "I want to be a founder, I have no expertise and a superficial understanding of the domain but I've got some tech, so here's a non-solution I'm selling as 'the best thing ever for you.'"

: calvinmorrison | 21 days ago
how about "this is illegal but with we have VC and an app it's now totally OK", see Uber's playbook.
: [OP] trubetskov | 21 days ago
i would say yes it does. such startups have no future. the case of one of the most famous ai-therapy-replacers Woebot proved this by its bancrupcy and $100M+ investment burnt.

jabiko | 21 days ago

Starting at page 20, the report looks like someone let AI create a HTML page and then just converted it to PDF. There are misplaced "Calendar", "Money Bag" and "Globe" emojis floating above the title of each startup analysis.

I suspect that nobody, including the authors, ever read this 391 page report.

: [OP] trubetskov | 21 days ago
You suspect wrong. We did this report for ourselves to correct the strategy and decided to make it public. All data is accurately checked, all numbers and maths can be easily checked by yourself.

cyanydeez | 21 days ago

The problem with mental health monetization is: it doesn't pay to make people feel better.

The problem with most of capitalism, is the same. There's just more money to be made from people with problems than solving problems.

: [OP] trubetskov | 21 days ago
not in every case. i think that we are making a very helpful tool - Mentalium - Voice CBT Diary. I made it for myself and i undrestand the problemts it solves really really good.

w10-1 | 21 days ago

Post-mortem's are very useful, though often not welcome.

I appreciate the top-line summary and the approach looks promising, but I don't have time to sort through the data or details.

Both are likely common, so don't be disheartened by complaints :)

: [OP] trubetskov | 21 days ago
We did this research for ourselves and started with "by-hand" inspection of every dead startup we could find. and i can say we found almost every. so you can spend no time reading every story and read only summary - it contains all our thoughts and this is the thing we used to make decisions on Mentalium's strategy.